Smart Auto Mapping Tool (SAMT)

| Overview

Smart Auto Mapping (SAMT) introduces an AI-powered automation layer within CCU that simplifies the process of mapping BACnet devices to Hayloft models.

This feature enables:

Automatic model suggestion
Intelligent point-level mapping
Seamless integration within CCU workflows

A screenshot of a computer application

AI-generated content may be incorrect.

By leveraging machine learning (BERT/OpenAI-based semantic similarity with cosine scoring), SAMT significantly reduces manual configuration effort and improves mapping accuracy across deployments.

| Objectives

Reduce manual effort (hours → minutes per device)
Improve mapping accuracy using ML-based similarity
Enable scalability across multi-site deployments
Streamline commissioning workflows
Demonstrate AI-assisted configuration (PLG enablement)

| Accessing Smart Auto Mapping Tool (SAMT)

The Smart Auto Mapping Tool (SAMT) is part of the custom equipment pairing workflow.

When pairing custom BACnet/Modbus equipment to the CCU.

There is a warning to Intialize the BACnet from Settings.

Once the BACnet is initialized, the BACnet model equipment configuration is made available.

The user is prompted to select a mode to map the points of the equipment being paired, as shown below.

Select IP configuration and click Start Auto Discovery to proceed with the AI -asssisted Smart Auto Mapping mode workflow.

Best Match Scenario

Select the connected device from the available list.

The tool auto-suggests matched models in terms of percentage for discovered BACnet devices as follows.

Good Match Scenario

Device

Suggested matches by the AI

Uses weighted similarity across:
- BACnet ID (highest weight)
- Object Name
- Object Type
- Units / Range / Enum
- Point Count similarity
Displays confidence score (%)
Auto-selects model if ≥ 90% match